基于对角递归神经网络的冷凝器污脏测量  

Measurement of fouling in condenser based on diagonal recurrent neural network

作　　者：樊绍胜[1] 王旭红[1] 王耀南[2] 

机构地区：[1]长沙理工大学电气与信息工程学院,湖南长沙410077 [2]湖南大学电气与信息工程学院,湖南长沙410082

出　　处：《传感器技术》2004年第8期58-60,共3页Journal of Transducer Technology

基　　金：国家自然科学基金资助项目(60075008)

摘　　要：提出了一种在线测量冷凝器污脏的新方法。该方法选取传热端差作为研究对象,运用对角递归神经网络建立变工况端差模型,成功分离出冷凝器污脏对端差的影响,准确实现冷凝器污脏程度的在线监测。根据此方法,研制了以数字信号处理器为核心的测量装置,并进行了现场试验,试验结果证明:该方法有效、可行。A novel method for measuring fouling in condenser is proposed.In the method,terminal temperature difference is chosen to reflect the fouling state,diagonal recurrent neural network is applied to separate the influences of both the fouling and off-design condition on terminal temperature difference.On-line measurement of the fouling in condenser is realized exectly.Based on it,an apparatus is developed and test on an actual condenser is conducted.The test result proves the proposed method to be effective.

关 键 词：冷凝器污脏 在线测量 传热端差 对角递归神经网络 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]